|Publisher||Society of Petroleum Engineers||Language||English|
|Content Type||Conference Paper|
|Title||Simulation of Gelled Polymer Treatments in the Arbuckle Formation, Kansas|
R. Barati, D.W. Green, and J.T. Liang, U. of Kansas
SPE/DOE Symposium on Improved Oil Recovery, 22-26 April 2006, Tulsa, Oklahoma, USA
|Copyright||2006. Society of Petroleum Engineers|
Gel polymer treatments have been carried out on a large number of producing wells in the Arbuckle Formation of Central Kansas. These treatments have been aimed both at reducing water coning from an underlying aquifer and increasing oil production. This paper describes computer simulation of two of these gel-treated wells. One is a well for which the treatment was successful and the second is a well which responded far less favorably. For each well, performance before, during and after a treatment was modeled using single porosity and dual permeability models, i.e. reservoir simulators that are based on matrix flow only or on fracture-matrix flow. For the fracture-based model, it was assumed that the producing well was connected to an underlying aquifer through a high vertical permeability fracture system.
Well performance during the polymer injection period of the treatments and the post-treatment behavior were such that there were major differences in history matching between the single porosity model (no fractures) and the fracture model. For the well that was treated successfully, behavior during polymer injection as well as post-treatment performance could be matched only with the dual permeability model. No reasonable set of parameters was identified for the single porosity (matrix flow) model that allowed a suitable history match of these same data. For the well that was treated unsuccessfully, the converse was true. An acceptable history match was obtained only with the single porosity model or with the fracture model with a very sparse fracture density. The results indicate that the presence of natural fractures connected to an aquifer is an important feature of the Arbuckle wells that respond favorably to polymer gel treatments.
The Arbuckle formation is the main oil producer in Kansas, responsible for approximately 36% of the total produced oil in Kansas. The geology of the Arbuckle is very complicated with a mixture of fracture, karst, and matrix-dominated architectural systems. This reservoir is supported by a strong bottom aquifer and water coning is the main problem in most of the Arbuckle wells. After many years of production, the water cut of most of the wells has increased to more than 98%.
To delay water coning, most of the wells in the Arbuckle were completed only in the top few feet of the productive interval. The reservoir is not well characterized due to the absence of field cores and the lack of welllog data for the full productive intervals. In most cases, the precise depths of the bottom of the productive intervals are not known.
Cr(III)-polyacrylamide gels have been used successfully to treat high water-cut wells in the Arbuckle. Most successful treatments resulted in an immediate decrease in the water production to less than half of the pre-treatment level and an increase in the oil production over the pre-treatment level. However, the water and oil production rates usually returned to the pre-treatment levels in less than one year. Some treatments were, however, unsuccessful where the gels had little or no effect on the oil and water rates.
The objective of this study is to determine the role of fracture versus matrix flow in the performance of gel treatments in the Arbuckle wells. Single-well models with radial coordinates were developed to simulate and compare the pre-treatment, treatment, and post-treatment flow behavior of a successful case (Hall B#4) and an unsuccessful case (McCord A#4). The CMG black oil simulator was chosen for our simulation study. Both the single porosity and dual permeability models were used to simulate the well performance. Conclusions were based on the quality of the history match using different models.
|File Size||210 KB||7|